• Proceedings of the Korean Society of Laser Processing Conference
• /
• 1998.11a
• /
• pp.0-0
• /
• 1998

-Part I- As market needs become more various, the production of smaller quantities of a wider variety of products becomes increasingly important. In addition, in order to meet demands for more efficient production, long-term unmanned factory operation is prevailing at a remarkable pace. Within this context, laser machines are gaining increasing popularity for use in applications such as cutting and welding metallic and ceramic materials. FANUC supplies four models of $CO_2$ laser oscillators with laser power ranging from 1.5㎾ to 6㎾ on an OEM basis to machine tool builders. However, FANUC has been requested to produce laser oscillators that allow more compact and lower-cost laser machines to be built. To meet such demands, FANUC has developed six models of Slab type YAG laser oscillators with output power ranging from 150W to 2㎾. These oscillators are designed mainly fur cutting and welding sheet metals. The oscillator has an exceptionally superior laser beam quality compared to conventional YAG laser oscillators, thus providing significantly improved machining capability. In addition, the laser beam of the oscillator can be efficiently transmitted through quartz optical fibers, enabling laser machines to be simplified and made more compact. This paper introduces the features of FANUC’s developed Slab type YAG laser oscillators and their applications. - Part II - All-solid-state lasers employing laser diodes (LD) as a source of pumping solid-state laser feature high efficiency, compactness, and high reliability. Thus, they are expected to provide a new generation of processing tools in various fields, especially in automobile and aircraft industries where great hopes are being placed on laser welding technology for steel plates and aluminum materials for which a significant growth in demand is expected. Also, in power plants, it is hoped that reliability and safety will be improved by using the laser welding technology. As in the above, the advent of high-power all-solid-state lasers may not only bring a great technological innovation to existing industry, but also create new industry. This is the background for this project, which has set its sights on the development of high-power, all-solid-state lasers with an average output of over 10㎾, an oscillation efficiency of over 20%, and a laser head volume of below 0.05㎥. FANUC Ltd. is responsible for the research and development of slab type lasers, and TOSHIBA Corp. far rod type lasers. By pumping slab type Nd: YAG crystal and by using quasi-continuous wave (QCW) type LD stacks, FANUC has already obtained an average output power of 1.7㎾, an optical conversion efficiency of 42%, and an electro-optical conversion efficiency of 16%. These conversion efficiencies are the best results the world has ever seen in the field of high-power all-solid-state lasers. TOSHIBA Corp. has also obtained an output power of 1.2㎾, an optical conversion efficiency of 30％, and an electro-optical conversion efficiency of 12%, by pumping the rod type Nd: YAG crystal by continuous wave (CW) type LD stacks. The laser power achieved by TOSHIBA Corp. is also a new world record in the field of rod type all-solid-state lasers. This report provides details of the above results and some information on future development plans.

• Aerospace Engineering and Technology
• /
• v.4 no.1
• /
• pp.108-113
• /
• 2005

Strucutral and hygrothermal analysis for a composite tube is carried out in this study, that provides critical parameters for the design of a highly dimensionally stable space telescope. Carpet plots for laminate effective engineering constants are generated and used for the best tube lay-ups with high elastic modulus and highly insensitive to thermal and moisture expansion, which is essential for maintaining optical alignment of opto-mechanical system under random force applied during a launch campaign and orbital thermal load. Despace in the longitudinal direction under hygrothermal load of the tubes constructed with the selected lay-ups is calculated for the validation of lay-up designs on the dimensionalstability. Dynamic analysis is also carried out to feature the resonant behaviour. A zig-zag triangular element accurately representing through thickness stress variations for laminated structures is developed in this study and incorporated into the structural and hygrothermal analysis.

• Journal of Institute of Control, Robotics and Systems
• /
• v.5 no.4
• /
• pp.419-426
• /
• 1999

A novel motion analysis system is presented in this paper. The proposed system is inspired by processing functions observed in the fly visual system, which detects changes in input light intensities, determines motion on both the local and the wide-field levels. The system has several differences from conventional motion analysis system. First, conventional systems usually focused on matching similar feature or optical flow, but neural network is applied in this system. Back propagation is used by learning method, and Tine Delay Neural Network (TDNN) is also used as analysis method. Second, while conventional systems usually limited on only two frames of sequence, the proposed system accept multiple frames of sequence. The experimental results showed a 94.7% correct rate with a speed of 71.47 milli seconds for real and synthetic images.

• Journal of Korea Multimedia Society
• /
• v.23 no.8
• /
• pp.977-985
• /
• 2020

Aiming at the problem of complex feature extraction and low accuracy in human action recognition, this paper proposed a network structure combining batch normalization algorithm with GoogLeNet network model. Applying Batch Normalization idea in the field of image classification to action recognition field, it improved the algorithm by normalizing the network input training sample by mini-batch. For convolutional network, RGB image was the spatial input, and stacked optical flows was the temporal input. Then, it fused the spatio-temporal networks to get the final action recognition result. It trained and evaluated the architecture on the standard video actions benchmarks of UCF101 and HMDB51, which achieved the accuracy of 93.42% and 67.82%. The results show that the improved convolutional neural network has a significant improvement in improving the recognition rate and has obvious advantages in action recognition.

• Proceedings of the IEEK Conference
• /
• 2000.09a
• /
• pp.837-840
• /
• 2000

동영상에서 움직이는 객체 분할 및 모션 예측을 동시에 수행할 수 있는 연구는 다양한 방법으로 시도 되어 왔다. 실제 이미지를 서로 다른 움직임이나 서로 다른 공간적인 특정 영역으로 분리 될 수 있다고 가정 한다면 복수의 객체 또는 객체의 움직임으로 표현 할 수 있다. 객체 분할 측면에서 볼 때 효율적인 분할을 위해서는 특징 입력 벡터의 선택이 중요한 변수로 작용한다. 본 연구에서는 정밀한 객체 분할을 위해 밝기, 질감(Texture) 정보와 같은 정지영상의 특징 입력 벡터와 움직임 벡터 같은 동영상의 특징 입력 벡터를 동시에 사용한다. 분리된 객체는 각각의 클래스를 구성하게 되고 이를 위한 클래스 분류기로서 Median Radial Basis 신경 회로망을 사용한다. 객체 분할과 움직임 예측을 위해서 확률적 방법을 통한 에너지 함수를 구하고 비용함수를 도입한다. 신경 회로망의 각 Basis 함수는 영상의 특정한 영역에서 활성화되며 객체의 분류를 위해 신경 회로망 출력으로 가중치의 합으로서 나타나게 된다.

• Journal of the Korean Ceramic Society
• /
• v.26 no.5
• /
• pp.645-654
• /
• 1989

Specimens of WC-Co were indented to measure the resulting crack size and unindented samples were fractured in 3-point flexure to obtain the strength and to measure characteristic features on the fracture surface. Fracture toughness was determined using fractography and compared to those determined using identation techniques. We show that principles of fracture mechanics can be applied WC-Co composites and can be used to analyze the fracture process. The fracture surfaces were examined by scanning electron microscopy and optical microscopy. Characteristic feature observed in glasses, single crystals and polycrystalline materials known as mirror, mist, hackle, and crack branching were identified for these composites. We discuss the importance of fracture surface analysis in determining the failure-initiating sources and the failure behaviorof WC-Co composites.

• Journal of The Korean Astronomical Society
• /
• v.47 no.4
• /
• pp.123-135
• /
• 2014

A silicate carbon star is a carbon star which shows circumstellar silicate dust features. We collect a sample of 44 silicate carbon stars from the literature and investigate the validity of the classification. For some objects, it is uncertain whether the central star is a carbon star. We confirm that 29 objects are verified silicate carbon stars. We classify the confirmed objects into three subclasses based on the evolution phase of the central star. To investigate the effect of the chemical transition phase from O to C, we use the radiative transfer models for the detached silicate dust shells. The spectral energy distributions and the infrared two-color diagrams of the silicate carbon stars are compared with the theoretical model results. For the chemical transition model without considering the effect of a disk, we find that the life time of the silicate feature is about 50 to 400 years depending on the initial dust optical depth.

• Proceedings of the KSME Conference
• /
• 2004.11a
• /
• pp.1510-1515
• /
• 2004

Aggregability of red blood cells (RBCs) was determined by a laser backscattering light analysis in a microfluidic channel. Available techniques for RBC aggregation often adopt a rotational Couette-flow using bob-and-cup system for disaggregating RBCs, which causes the system to be complex and expensive. A disposable microfluidic channel and vibration generating mechanism were used in the proposed new detection system for RBC aggregation. Prior to measurement, RBC aggregates in a blood sample were completely disaggregated by applying vibration-induced shear. With the present apparatus, the aggregation indexes of RBCs can be easily measured with small quantities of blood sample. The measurements with the present aggregometer were compared with those of LORCA and showed a strong correlation between them. The aggregability of the defibrinogenated blood RBCs is markedly lower than that of the normal RBCs. The noble feature of this design is the vibration-induced disaggregation mechanism, which enables to incorporate disposable element that holds the blood sample.

• Proceedings of the KSRS Conference
• /
• 2005.10a
• /
• pp.283-285
• /
• 2005

To account for the importance of variable in multi-sensor data fusion, random forests are applied to supervised land-cover classification. The random forests approach is a non-parametric ensemble classifier based on CART-like trees. Its distinguished feature is that the importance of variable can be estimated by randomly permuting the variable of interest in all the out-of-bag samples for each classifier. Supervised classification with a multi-sensor remote sensing data set including optical and polarimetric SAR data was carried out to illustrate the applicability of random forests. From the experimental result, the random forests approach could extract important variables or bands for land-cover discrimination and showed good performance, as compared with other non-parametric data fusion algorithms.

• Journal of the Korean Society of Manufacturing Process Engineers
• /
• v.12 no.3
• /
• pp.61-68
• /
• 2013

Automatic measurement system is required since cycle time and cost of production are increased by various factors in manual systems. This paper presents a machine vision based prototype measurement system for the automotive axisymmetric shape parts which are generally measured by a manual system that is required the tolerance of the part is very small on each machined surface. This measurement system adopts a method in which optical lens is transferred along the profile of the part to minimize measurement cycle time. The main interest of this paper is a development of an optimum measurement algorithm to the outside diameter of the parts that can be applied to various combinations of hardware. The operating system used to implement the whole system is Window XP and corresponding environment.